Minturnae – porticus with tabernae in front of the ‘Macellum

Commerce and the senses: everyday work and the Roman urban landscape #RACTRAC16

This is the full text of the conference paper that I wrote for the session on ‘Sensing Rome’ at the Roman Archaeology Conference in Rome, 19 March 2016. 

There is no doubt that sensory perceptions are fundamental to our understanding of urban landscapes – present and historical – and, as the other papers have already made clear, movement is a central component of the urban sensory experience. If we want to understand what Roman urban landscapes ‘were like’, the smells, sounds, sights and feels that one could encounter whilst moving through the city cannot be left out of the picture, nor – it has to be added –is their position in the daily or seasonal rhythm, or indeed their place in town: the sensory encounters in the outskirts of town may be fundamentally different from those of the city centre, and the summer experience may be dramatically different from the winter experience – not only in terms of perceptions to be encountered, but also in the sheer number of encounters, because of varying patterns in movement. It is also true that sensory experiences matter to people in motion: encountered unexpectedly, they may contribute to spontaneous, in-motion decision-making, such as the maybe often half-conscious decision to look where the sound or smell perceived is actually coming from, or the much more conscious decision to halt. Alternatively, people, on their way from a random point A to a random point B, may adapt their journey because of expected sensory encounters—those that are to be avoided, or, perhaps more often those that are to be sought after.

Pompeii – bakery VI 2, 6

This paper discusses sensory experiences related to commerce, focusing particularly on the urban landscape of Pompeii, where the evidence for retail and manufacturing is most explicit. Commercial processes have of course the potential to contribute all kinds of sights, smells and sounds to urban sensescapes, particularly when activities take place in locations with close links to public space, and even more so if that public space attracted a lot of people in motion because it was central to the urban movement network, or because it was close to a highly frequented destination. Depending on its position in the urban landscape, commerce may have a larger or a smaller impact on the urban sensory experience, though it is of course essential to distinguish between the hours of the day in which commerce took place, and those hours in which it generally did not. Obviously, this paper will be mostly preoccupied with the commercially most viable hours of the day, but it should be noted that the sensory impact of commerce may extend beyond these: closed shops can make a visual impact, too, and while inactivity suggests a certain degree of silence, some of the smells may hang around for quite some time after the working day is over, particularly when not all of the waste has been cleaned up—if there is literary evidence suggesting shop holders could leave empty amphorae in front of their shop on the street, they may have left other stuff as well.

Urban Commercial landscapes in Roman Italy

Alba Fucens
Alba Fucens – taberna with counter on the Via dei Pilastri

It may be good to start with highlighting some key basic properties of urban space in Roman Italy, and of the way commerce has been embedded in it. First, it is important to emphasize the fundamental openness of urban space in Roman Italy, particularly when it comes to anything related to commerce and retail. Streets in cities like Pompeii and Ostia were surrounded by tabernae that were characterized by a wide opening that fostered interaction between the street and the inside. Tabernae could host all thinkable forms of retail, and a great deal of small-scale crafts too. The evidence from Pompeii leaves no doubt that they also did so: all major workshop types that are known at Pompeii, with the exception of tanneries, are attested in tabernae. Cities elsewhere in Roman Italy, as far as we know them, tend to present a similar picture: long rows of tabernae surrounded the main streets in the city centre.

However, it is relevant to emphasize that this picture is by no means the norm in the ancient Mediterranean. Greek cities, especially before the Hellenistic period, had few taberna-like facilities: the ‘shops’ that have been identified at sites like Olynthos and Priene were very few in number, but more importantly, they were very small and only had a rather narrow opening of the size of a normal door. The size and the lack of light makes it less likely that a lot of work was being done on their inside, and in any case made its sensory impact on people in motion incomparably lower. Even at Delos, which developed perhaps the most strongly commercialized landscape of the Hellenistic world that we know of, shops interact with the street in a way that is not comparable to what you see at Pompeii. In most parts of the Roman empire, tabernae never became as dominant as they were in Roman Italy, and few of the forms of investment in tabernae that can be identified at Pompeii, Ostia and Rome are also attested in, for instance, Roman Africa, Spain or Gaul. One logical explanation for this is that in places like Ostia and Pompeii, which were right in the wealthy heart of the empire, imperial power brought imperial wealth, and created a level of urban consumer demand that elsewhere was much lower. In Pompeii, there also is a clear chronological correlation between the emergence of Rome as a pan-Mediterranean power in the second century BC, and the commercialization of the urban landscape. In a way, the sensory impact of commerce on urban landscapes was a fruit of the empire, and of the prosperity that Rome’s political power brought to Italian urban communities.

Pompeii – map of the city highlighting the tabernae
Pompeii – map of the city highlighting the tabernae

Yet, this is not the whole story: there is a cultural component to the proliferation of tabernae as well, and I would argue—contrary to the tendencies of some past scholarship and despite the dismissive comments by some Roman authors who are not all called Cicero—that Roman culture, at least in practice, was mostly liberal towards commerce, especially if you compare it with some other historical periods. This is true on the microscale as well as on the macroscale. On the macroscale it should be noted that, even though Roman governments were actively involved in building commercial facilities, and thus played a key role in shaping Roman commercial landscapes, they never appear to have done so in a very prohibitive way—contrary to, e.g. the imperial cities of ancient China, which had designated areas for commercial activity, or the cities of the medieval Islamic world, where a lot of commercial activity ended up concentrated in Bazaars and Souqs. In Roman Italy, anyone who wished to could build shops or workshops, basically, wherever he or she wanted, with the result that many people did this: at Pompeii, all classes of urban house-owners were involved in retail and manufacturing. This led to rather organically grown commercial landscapes that clustered around the busier roads of the city and were as dense and vast as local consumer demand could carry. It is probably anachronistic to claim that Roman commercial landscapes were mostly created ‘by the market’, but there certainly appears to be a clear economic logic to their shape. At the same time, as I have argued several times in the past, there is very little evidence suggesting that, on the micro-scale, production processes as a rule were consciously isolated from the public urban environment: in retail and manufacturing, you could do whatever you wanted, wherever you wanted, with very few exceptions.

The smell of commerce

Yet that Romans had a mostly liberal approach to the sensory impact of crafts and retail does not mean that the resulting sensory landscape was not in some way or another spatially articulated. In what follows I will use the commercial landscape of Pompeii to discuss how things worked out in practice. I will particularly focus on two forms of impact that are to some extent visible in the archaeological record: heat and smell. Noise, of course, is also relevant, but it is much harder to find indications of places where, for instance, a lot of hammering took place. The visual impact of most tabernae is impossible to reconstruct as many were found devoid of any indication as to their use. Arguably, while in many cases the walls were lightly coloured, the inner parts of tabernae would not easily have visual focus, as the relative lack of light compared to the street or the portico in front of them obscured what happened inside except for those actively looking. This is a bit easier with heat and smell, though it is almost impossible to build up a complete picture as there are too many unknowns.

Hot food

Pompeii – taberna VI 3, 19-20, with counter and heating installation
Pompeii – taberna VI 3, 19-20, with counter and heating installation

There is a number of processes of retail and manufacturing that spread heat or smell. First and foremost, there is the retail of hot food, which, as Steven Ellis and others have argued, can be identified in the archaeological record through the taberna counters including the remains of a heating facility. Evidence for hot food retail has been found in great numbers, with more than 120 examples securely identifiable: about one eighth of Pompeian tabernae seems to have sold hot food. In our most positive population scenario for Pompeii this comes down to one known food retail outlet for every 100 people. Probably, they did not all do precisely the same thing, but they all had the heating installation which is extremely likely to have produced smells.

Pompeii – map of the city highlighting the tabernae. Green: tabernae with installations suggesting hot food retail
Pompeii – map of the city highlighting the tabernae. Green: tabernae with installations suggesting hot food retail

If you plot the 127 tabernae with cooking installations in the catalogue of Ellis on a map, it also is clear that they are rather evenly spread over Pompeii’s urban topography, with several minor concentrations in the environment of large public complexes, but no truly specialized food districts. As a consequence, the smell of fresh food being prepared would have been a characteristic element of the entire Pompeian urban landscape—wherever you went, chances were high that you would encounter one or more of these food shops on your way. Particularly, it should be noted that food shops were also well-represented in regions with otherwise little or no commercial activity, such as in region 1 south of the Via dell’Abbondanza, where almost all shops were involved in hot food retail, and up in region six, where there is a similar picture. Indeed, along the more densely commercialized through-roads, the proportion of tabernae serving hot food is relatively low, especially in the city centre. Nevertheless, during business hours, the smell of hot food would have been almost inevitable at Pompeii.


Pompeii – taberna IX 6, a.1: fulling installations (Cleaning the Laundries, 2006).

Besides food retail, there also is a number of recognizable workshops that used heating installations or potentially smelly chemicals. One of these, of course, is the fullonica, though I have argued at several places that the smell of fulling has been overestimated in the past. In any case, fulling was freely exposed to the urban environment, and there is nowhere evidence for measures aimed at reducing smell. Dyeing workshops, comparatively, have a much clearer environmental impact because of their use of heated cauldrons; though the shape of the cauldrons suggests that they could in fact be covered, they produced heat and smoke. It is unclear what was cooked in the bowl-shaped furnaces of the lanifricariae, but even if it was just raw wool (which I doubt), the process could have made considerable impact on passers-by in places where the workshop had a direct connection to the street. Finally, there are the forty-odd bakeries which produced heat, smoke and the smell of freshly baked bread, though in practice, it often was mixed by the smell of the animals operating the mills.

Pompeii – lanifricaria VII 12, 17.21
Pompeii – lanifricaria VII 12, 17.21

About half of the workshops of these four types had a direct connection with the street in a way that theoretically made it possible for people to smell what was going on. If we plot these workshops on a map, it is immediately clear that they are much less broadly spread over the urban topography than the tabernae selling hot food. They are more or less concentrated in the zone north and east of the forum, and there were few publicly visible workshops in the east of the city. Yet more than the macro-picture, it is the micro-picture that matters: work rooms, unless they were situated in a taberna, generally opened on streets that played a more marginal role in the urban street system. For instance, the entrances to the work rooms of the lanifricariae in the heart of region VII were all concentrated on the marginal Vicolo del Balcone Pensile and Vicolo della Maschera rather than on the much more densely commercialized Vicolo di Eumachia. Thus, while in principle the greatest variation in sensory stimuli was to be found in the zone around the monumental city center, the major through-roads were mostly free from smells and heat caused by artisanal activity. This holds true for Via della Fortuna,  the east stretch of Via dell’Abbondanza, Via degli Augustali and via del Foro, as well as for the central stretch of Via Stabiana.

Pompeii – map of the city highlighting the tabernae and the workshops with a direct connection to the street Red: bakeries. Purple: lanifricariae. Blue: fullonicae. Yellow: dyeing workshops.
Pompeii – map of the city highlighting the tabernae and the workshops with a direct connection to the street. Red: bakeries. Purple: lanifricariae. Blue: fullonicae. Yellow: dyeing workshops.

The marginality of smell?

If we combine the workshops and the tabernae selling hot food onto one map, the overall picture is that the most densely commercialized roads have relatively low amounts of establishments producing smells and heat, even though hot food retail is slightly more common on the central thoroughfares than manufacturing. The question is how this is to be understood.

Pompeii – map of the city highlighting the tabernae and the workshops with a direct connection to the street Green: hot food outlets. Red: bakeries. Purple: lanifricariae. Blue: fullonicae. Yellow: dyeing workshops.
Pompeii – map of the city highlighting the tabernae and the workshops with a direct connection to the street. Green: hot food outlets. Red: bakeries. Purple: lanifricariae. Blue: fullonicae. Yellow: dyeing workshops.

The absence of crafts and food retail has, of course, been noted for west stretch of Via dell’Abbondanza by Wallace Hadrill who suggested, two decades ago, that there was a ban on certain banausic types of commerce along this road because of its role as a processional way. Yet there is room for a different argument, if one starts with the observation that the situation in Via degli Augustali, Via del Foro, large parts of Via della Fortuna, and indeed in Via Stabiana too, is not fundamentally different, though there are just enough tabernae serving hot food making government policy an unlikely explanation. This suggests that it was not top-down policy that shaped this landscape, but bottom-up strategy: people along these roads simply decided to use their tabernae for other purposes than selling hot food.

Understanding the commercial topography

The question is why they did so. Actually, the most straightforward clue for an answer comes from the buildings themselves. It is striking that, just three or four of the more than eighty tabernae associated with major public buildings—the baths, the macellum—contained a workshop or a facility to prepare hot food. The picture with the tabernae associated with the largest houses of the city is essentially similar. Indeed, while the largest houses of Pompeii included a disproportional number of the city’s tabernae, a disproportional number of hot food shops was associated with tabernae related to smaller houses. The large majority of the tabernae with facilities for food preparation along the Via della Fortuna and the Via Stabiana belong to buildings that, on the grounds of their dimensions, belong to the bottom half of Pompeii’s urban hierarchy. Hence, the retail of hot food is concentrated in places where there are many small buildings, such as in insula VII 3, and in insula IX 1.


Economically, this also makes sense: while selling food may make it possible to make ends meet, it is not awfully lucrative, and especially in the city centre, other forms of retail were more attractive for those who had to decide about the fate of a taberna. In other words: not cultural preferences or norms dictated the shape of Pompeii’s commercial topography, but a differentiation of investment patterns between the economically more and less powerful, and, one has to add, the dimensions of the plots laid out when insulae were first built up, several centuries BCE. By coincidence, this led to a relative marginalization of hot food facilities, which therefore cluster around the fringes of the city centre, in places where competition with more lucrative forms of commerce was less tense, and to zones with many small houses.


To conclude, it has been argued, recently, that Roman cities had a ‘taberna economy’, and while I have my doubts as to whether this is an accurate term to capture the complexities of urban economies throughout the Roman world, the spatial impact of tabernae on urban landscapes in Roman Italy is undeniable. What we see in Pompeii is very much an ‘in your face economy’, and this is likely to have been echoed throughout Roman Italy. By consequence, moving through the cities of Roman Italy meant moving through an urban world that was heavily commercialized, and where all kinds of activities could be going on at a short distance of each other. At the right hours of the day, few things potentially were as central to the sensory experience of people in motion as what was going on in the tabernae, especially because, for practical reasons, tabernae could only function when their main door opening was open—otherwise, the room would be too dark. This was different for workshops, but street entrances to work rooms were, as we have just seen, generally concentrated in less busy streets anyway. Yet I have tried to argue that, at least in the one place where we can actually do more than just map tabernae, there was a logic to the chaos, and one that may have resulted in a relatively less smelly and noisy character of the main commercial areas—for economic reasons. This does not mean that there was no variation, rather that the most important sense as far as commerce was concerned may often have been sight. This does have implications.

Pompeii – Via Stabiana, between Stabian Baths and insula IX 1, looking north.

For many of the pedestrians in the streets of Pompeii, the sequence of shops would be familiar—they walked their walks, knew their daily environment, and most of the time may simply have taken the visual landscape for granted because it was too familiar, or because the weather of the day had a stronger impact on their experience than the usual people doing their usual things in their usual place. For those who were moving through less well-known territory, the nature of the experience would often be instantly understandable, and many of the scenes they saw would have been recognizable. Moreover, the natural sensory focus of people in motion is not undirected, and therefore is limited: people are navigating, and have to look in front of them to see where they are going, and which obstacles they may encounter on their way, which generally would have meant that they were not looking towards the inside of tabernae, as this picture makes clear—often, people may have needed a trigger that would make them take a look. In a way thus, commercial live, however visible it was potentially, in practice often may have remained rather on the background, unless people were actively looking for something.